Symposium on Advances in Approximate Bayesian Inference

December 2, 2018
Le 1000 Conference Center
1000 Rue de la Gauchetière Ouest
Montréal, QC H3B 0A2, Canada


Registration

7:30 - 8:30 Registration (Coffee will be served)

Session 1

8:30 - 8:40 Introduction
8:40 - 9:10 Invited Jean-Michel Marin
9:10 - 9:30 Contributed Maria I. Gorinova: Automatic Reparameterisation in Probabilistic Programming
9:30 - 10:00 Invited Tamara Broderick
10:00 - 11:00 Coffee Break and Poster Session (Paper IDs 1-24)

Session 2

11:00 - 11:20 Contributed George Tucker: Doubly Reparameterized Gradient Estimators for Monte Carlo Objectives
11:20 - 11:50 Invited Tom Rainforth
11:50 - 12:10 Contributed David Burt: Explicit Rates of Convergence for Sparse Variational Inference in Gaussian Process Regression
12:10 - 13:40 Lunch Break (on your own)

Session 3

13:40 - 14:10 Invited Thomas Schon
14:10 - 14:30 Contributed Badr-Eddine Chérief-Abdellatif: Consistency of ELBO maximization for model selection
14:30 - 15:00 Invited Sebastian Nowozin
15:00 - 16:00 Coffee Break and Poster Session (Paper IDs 25-45)

Session 4

16:00 - 16:20 Contributed Will Grathwohl: Scalable Reversible Generative Models with Free-form Continuous Dynamics
16:20 - 16:50 Invited Emityaz Khan
16:50 - 17:10 Contributed Matthew Hoffman: NeuTra-lizing Bad Geometry in Hamiltonian Monte Carlo Using Neural Transport
17:10 - 18:00 Panel
Alexander Alemi, David Duvenaud, Eric Xing, Ole Winther, Tamara Broderick
Moderator: Yingzhen Li